Social networks and symptomatology in recently hospitalized individuals with schizophrenia-spectrum disorders: A six-month longitudinal study

Individuals with psychosis have reduced social networks; however, little is known about whether social networks change over time, particularly after an acute episode requiring hospitalization. Furthermore, longitudinal associations between social network engagement, symptom severity, and the impact of illness duration remains unclear. Using linear mixed models controlling for age and sex, we analyzed self-reported social network engagement, positive and negative symptom severity, and self-reported paranoia in individuals with a schizophrenia-spectrum disorder (SSD; N = 68) during six months following psychiatric hospitalization and compared them with non-clinical participants (NC; N = 70). SSD participants reported significantly lower social network scores than NC participants at every time point (ps <.001), and significantly greater disruption in friendship than family ties (p = .012). Social networks remained stable over time in both groups (ps >.05), regardless of illness stage in the SSD group (ps >.05). Notably, lower social engagement with family at baseline predicted greater increases in paranoia over time (pcorrected = .048). Conversely, more severe paranoia at baseline predicted greater reductions in friend networks over time (pcorrected < .001). Finally, social network related to negative symptom severity at baseline (pcorrected =.01) and longitudinally (ps<.02). These findings suggest that clinical severity may influence the social network disruptions commonly found in psychotic disorders, especially during vulnerable periods of recovery following a psychotic exacerbation. Ultimately, this highlights the need for tailored post-discharge interventions that address social network deficits, particularly within families and in the context of severe paranoia.